Intersectional Fairness in Reinforcement Learning with Large State and Constraint Spaces

TL;DR

This paper addresses multi-objective reinforcement learning focused on fairness across intersecting demographic groups, proposing efficient algorithms for large state and constraint spaces, validated through experiments on complex graph MDPs.

Contribution

It introduces oracle-efficient algorithms for multi-objective RL with exponentially many objectives, extending fairness optimization to large and structured MDPs.

Findings

Algorithms efficiently solve multi-objective RL with many objectives.

Theoretical guarantees for large state and constraint spaces.

Experimental validation on graph MDPs demonstrates practical applicability.

Abstract

In traditional reinforcement learning (RL), the learner aims to solve a single objective optimization problem: find the policy that maximizes expected reward. However, in many real-world settings, it is important to optimize over multiple objectives simultaneously. For example, when we are interested in fairness, states might have feature annotations corresponding to multiple (intersecting) demographic groups to whom reward accrues, and our goal might be to maximize the reward of the group receiving the minimal reward. In this work, we consider a multi-objective optimization problem in which each objective is defined by a state-based reweighting of a single scalar reward function. This generalizes the problem of maximizing the reward of the minimum reward group. We provide oracle-efficient algorithms to solve these multi-objective RL problems even when the number of objectives is exponentially large-for tabular MDPs, as well as for large MDPs when the group functions have additional structure. Finally, we experimentally validate our theoretical results and demonstrate applications on a preferential attachment graph MDP.